| In recent years,with the continuous increase of car ownership,its safety problem has become increasingly prominent.As the core of automotive active safety,the perception system can effectively reduce the probability of accidents.Millimeter-wave radar is an important sensor in automotive sensing system because of its small size and little influence of weather conditions.In this paper,millimeter-wave radar is used as a sensor to detect obstacles for vehicles.The vehicle-mounted millimeter-wave radar system is designed.The distance detection,speed detection,threshold setting,dynamic and static target separation algorithm are studied,and MATLAB is used to realize above algorithm.The experimental platform is built to verify the system performance.The specific research content is as follows.1)This paper investigates the obstacle detection system of unmanned vehicle.The implementation methods of vehicle-mounted sensing system domestic and overseas are analyzed and compared.On this basis,the research status of vehicle-mounted millimeter-wave radar is analyzed.The design idea of obstacle detection system for unmanned vehicle based on millimeter-wave radar and the research content of this paper are proposed.2)The vehicle-mounted millimeter-wave radar system is designed,and the function and realization method of each module of the system are introduced in detail.On the basis of studying the principle of vehicle-mounted millimeter-wave radar,aiming at the frequency-modulated continuous wave(FMCW)system,which is the most widely used in vehicle-mounted millimeter-wave radar,the method of measuring target distance,velocity,Angle and other information using sawtooth wave is studied.3)The digital signal processing algorithm of the vehicle-mounted millimeter-wave radar is designed and realized by MATLAB.In order to solve the problem of acquiring the target distance information,the windowed 1-D FFT is used to process the intermediate frequency signal of the radar to extract its spectrum information.Constant false alarm rate(CFAR)algorithm is adopted to automatically adjust the detection threshold,and it improves the accuracy of obtaining the peak frequency of intermediate frequency signal.Combining the corresponding relation between peak frequency and target distance,the relative distance between measured target and radar is calculated.In order to obtain the velocity information of the target,the 2-d FFT method is used to process the radar intermediate frequency signal to obtain its initial phase information.The coherent phase difference method is used to reduce the influence of static target or noise in 1-D FFT results.Combining the corresponding relation between the phase and the target velocity,the velocity information of the measured target is calculated.Finally,for the problem of moving target recognition,density-based spatial clustering of applications with noise(DBSCAN)algorithm is used to cluster the detected lattice signals,and the detection results are classified according to different targets.The signal processing of automotive millimeter-wave radar is completed4)The host computer software program of millimeter-wave radar system is designed and completed.The host computer software can be connected with the millimeter wave radar hardware through a serial port.Radar parameters can be configured and the detection results can be displayed,saved and played back in real time.This paper introduces the overall structure of the host computer software and the functions of each module as well as the programming method of the software.An experimental platform is set up to verify the design of millimeter wave radar system.In view of the common scenes of automobile,the performance indexes of millimeter wave radar,such as the measurement accuracy and the maximum measurable range,are tested respectively in the indoor parking lot and the outdoor road.The feasibility and accuracy of the system and algorithm are verified... |
PDF Full Text Request